🚀 go-pugleaf

Using numerical data allows us to take a large quantity of data and reduce
it to one number or at least a small enough quantity of numbers.  Several
thousand, several hundred, or even several dozen data points are just too
much for an individual's brain to process.  Working with subjective data
does not have this advantage.  Is it still valid?  Yes.  Is it still useful?
Yes.  Can we process large quantities of it to obtain the true message
imbedded in it.  That is difficult.  How then can we evaluate suppliers in
such a way to get objective data?  Here are some statistics/questions that
would allow that.
*       Average delivery time
*       Average Cost
*       Rate on a scale of 1 to 5.  You can apply a verbal evaluation
criteria to each number.  For example:  1) fails to meet requirements in a
major way;  2)  fails to meet requirements in a minor way;  3) meets
requirements;  4) exceeds requirements in minor way;  5) exceeds
requirements in a major way.
*       Willingness to meet special needs
*       Flexibility
*       Effectiveness of their quality system
*       Effectiveness in meeting any other need you may have.
Note that by using this rating scale, you take a subjective response and
convert it to a number.  That number may not be as precise as some numerical
data, but it should still be a measure of how effective that company is.

I have assumed that all candidates under consideration all use standards
traceable to NIST and have a quality system.  Do the have internal audits
and a management review?  Will they allow  you to see those reports?  An
easy way would be to give preference to those that are ISO 9000 or ISO 25
certified.

Dave Loebach
OSHA
NRTL Program
david.loebach@cincinnati-tc.osha.gov

         >Frank Isackson said, "Everything that can be measured or described
in
         >words can be quantified.
         >[snip - sorry 'bout that]

         >Is it acceptable to rate suppliers using either only subjective
data
         >or a  combination of subjective and objective data? In other words
is any method that
         >does not rely exclusively on objective data a fundamentally flawed
method?  [typos
         >fixed correctly, I hope]

        > Dan O'Leary CQA, CQE, CQM, CRE
         >Business Process Executive for Manufacturing
         >doleary@qbe.picker.com